‘It would be smart to discuss this on telegram’: China’s digital territorialization project and its spatial effects on contentious politics

ABSTRACT Despite extensive literature examining digital activism, there is limited geographical research exploring the spatialities of digitally centred contentious politics. This article makes use of recent literature examining territory, terrain, sovereignty and digital territorialization projects to produce a conceptual framework that can make sense of the complex geographies of digital activism in contemporary China. It then uses this conceptual framework to analyse the spatial effects of digital territorialization projects and the development of digital sovereignty on those practising contentious politics in China. To do this, it uses the case study of #BeijingSurgery#, a Beijing-centred hashtag and instant messaging project attempting to contest narratives around the eviction of migrants from Beijing in 2017.


INTRODUCTION
In this article I build on recent scholarship examining territory to consider the forms that digital territory currently takes and can take (Elden, 2017(Elden, , 2021. To help in this process I make use of Möllers's (2021) concept of 'digital territorialization projects', originally used to examine the German experience, to understand how, since the mid-1990s, China has produced a complex, bordered and multi-scalar digital territory. In doing this I am interested in understanding what form Chinese digital territory takes, how China's digital territory is governed and how people within China's digital territory connect to other digital territories as well as the rest of the world. In extending Möllers's arguments I make a secondary argument, that the digital territory of the United States is, at the time of writing, the most populous digital territory in the world. By exploring US digital territory alongside Chinese digital territory I highlight the role that state and non-state actors play in digital territorialization projects, how territory can be used as a technology to achieve sovereignty over both space and people (Morris, 2022b), and why, due to data and digital colonialism (Couldry & Mejias, 2019;Kwet, 2019), it is strategically advantageous for the United States to stay out of debates on territory and sovereignty (Couture & Toupin, 2019). Through a case study of digital activism in China, the eviction mapping project #BeijingSurgery#, I highlight the importance of understanding digital territorialization projects globally, as when activists leave China's digital territory they often make strategic use of spaces in US digital territory, such as Instagram and Twitter, or multi-territorial digital spaces, such as Telegram, with data centres in at least the United States, the Netherlands and Singapore.
I will begin this exploration of digital territory's (re)production and the development of digital sovereignty from a relatively top-down position: the writing of Binxing Fang (Fang, 2018), the lead engineer of China's digital border, the so-called Great Firewall of China (Griffiths, 2019;Roberts, 2018), and an influential state-affiliated academic. Fang's academic writing is chosen here as it offers insights into how digital socio-spatial relations are understood at the highest levels of Chinese politics. I will build on this analysis by examining Chinese government documents, laws and regulationsincluding those used by Fang in their analysisto explore how digital territory is articulated legally and how sovereignty over digital territory is developing (Creemers, 2020). In doing this I note a legal turn in the Chinese digital territorialization project, a turn that enables the Chinese state to both govern internally and to act as a norm entrepreneur.
In the empirical sections of this article, through the #BeijingSurgery# case study, I provide a bottom-up perspective on digital territorialization projects and their spatial effects. #BeijingSur-gery# was an anti-eviction mapping project that contested the treatment of migrants in Beijing following a deadly fire in Xinjian Village in November 2017. #BeijingSurgery# started as a hashtag focused project in the days following the fire. The project was organized on Weixin, 1 where #BeijingSurgery#'s administrators-to-be discussed starting a hashtag project to map the ongoing evictions. Following this, a long-image 2 was collaboratively made using an online text editing site before this long-image was shared in a Weibo 3 original post (OP) along with the text #Beijing-Surgery#. This moment of posting created the OP, which could be reposted, liked or commented on, becoming a dynamic digital place; the hashtag, a larger digital place where multiple networked conversations could be had, and where people gathered for information and to create connections; and the project itself, an affective, networked public including project leaders, project members and general viewers who conversed within the projects semi-public hashtag (Boyd, 2011;Papacharissi, 2015). Unfortunately, after rising to the top of Weibo's hashtag metrics, after fewer than 20 hours, the #BeijingSurgery# hashtag was deleted resulting in the project's leaders and users being displaced (Morris, 2022a). Fortunately, the project had a backup Weixin instant messaging group. It was here that the project continued, though with significant limitations.
In analysing #BeijingSurgery#, I highlight four spatial effects of China's digital territorialization project on contentious political projects: deletion and displacement, the use of semi-private rather than semi-public spaces for gathering, the use of non-Chinese digital territories for activist practice, and mobility across a multitude of digital borders. In the discussion, I highlight what #BeijingSurgery#'s experience teaches us about digital activism more broadly, including the role of multiple territories in digital activism, and how digital activists may be simultaneously situated across multiple territories to balance out risks associated with their 'socio-spatial positionality' (Leitner & Sziarto, 2008, p. 158). I conclude the article by arguing that nobody need be bound by Sino-American imaginaries of digital territory or sovereignty and that alternative models of digital territory are possible. In doing this I encourage readers to support digital territories where projects such as #BeijingSurgery# can flourish, and not be silenced.

DIGITAL TERRITORIAL FRAMEWORK
In understanding how a digital territory emerges, the development of digital sovereignty and the production of a digital terrain, this article builds on the scholarship of Stuart Elden, who argues that territory is a 'political technology' (Elden, 2010, p. 801) that 'should be understood not as a simple bounded area, but in multiple registers' (Elden, 2017, p. 8). For Elden, territory is not a single political technology but rather 'a bundle of political technologies' (Elden, 2013, p. 322), and it is not a universal but something that is (re)produced across time and space by numerous actors, from the state and its various appendages to everyday people who live, work and exist around borders of multiple kinds. As a political technology, territory can be used to dominate voluminous space and the people inhabiting it (Billé, 2020), and there are ongoing international negotiations over how territory can be produced underground and undersea, on land, in the sky, in space and in the poles (Adey, 2010;Campling & Colás, 2018;Gordillo, 2018;Starosielski, 2015).
As a political technology through which power over people is practiced, territory and territorialization has an important influence over hybrid activist geographies (Castells, 2015;Halvorsen, 2021), including the roles that both human and non-human actors, such as algorithms, play in such geographies. States are engaged in active legal, infrastructural, commercial and moral contestation over which actors can use digital spaces, which states control these spaces, what institutions and people have access to these spaces, and how the by-products of digital interactions are used (Couldry & Mejias, 2019;Creemers, 2020;Kwet, 2019;Starosielski, 2015). As more social, political, cultural, economic, scientific and militaristic actions take place in digital spaces there is an increasing drive for governments and institutions to carve out spaces where they are in control. To do this, states are actively engaged in practices that border digital spaces, and while many states, governments and institutions refrain from using the potentially loaded language of 'territory', 'border' and 'sovereignty', there are also a large number of states and corporations using digital territory as a political, economic, cultural and militaristic technology (Barnard-Wills & Ashenden, 2012;Howard et al., 2011;Kaiser, 2015).
The role of territory in digital relations is particularly visible in Norma Möllers's (Möllers, 2021) recent scholarship, which examines how digital territory is produced through 'territorialization projects'. Exploring the German model, Möllers describes how: states mobilize scientists and engineers in order to transform globally distributed information infrastructure into bounded national territory, and how they invest it with patriotic meaning, thereby making 'digital territory.' 'Digital territory,' in other words, is nationalized information infrastructure. (p. 114) For Möllers, digital territory is 'nationalized as in materially under state control and nationalized as in invested with normative ideas about nation and citizenship' (p. 116). But, as Morris (2022a) suggests in their work on Chinese digital geographies, both non-state and state actors play key roles in the moral and material dimensions of digital territorialization. This means that nationalization is just one articulation of the moral dimensions of digital territorialization, and a variety of moral articulations can potentially enable digital territorialization projects to begin. Looking at the United States, for instance, it becomes possible to see a loosely regulated and bordered digital territorywhat Bratton (2018) describes as a 'hemispherical stack' built around Google, Apple, Facebook and Amazonwhere private corporations play an important role in the expansion of US digital power.
Extending these spatial conceptualizations of digital relations, and influenced by Jessop et al. (2008), in this article I put forward a framework that helps highlight the complex geographies and power relations involved in digital relations. Building on the above discussion, in this article digital territory is understood as a political technology through which institutions, most notably nation-states and international organizations, attempt to gain power over digital spaces, digital places, users, the companies operating digital spaces, and digital infrastructure through a range of bordering and territorialization practices. Digital spaces are distinct digital services hosted on servers which can be accessed by humans and non-humans in ways that enable a social life, broadly defined, to emerge within the coded environments (Amoore & Raley, 2017;Kotliar, 2020;Massey, 2005). They are 'a constellation of processes rather than a thing' (Massey, 2005, p. 141) and they can include multiple digital places, with examples including Weixin, Facebook, Bilibili and Telegram. Sitting within digital spaces, digital places are distinct assemblages that emerge from the interaction between humans and non-humans within digital spaces. Digital places can take many forms, with examples including instant messaging groups, hashtags, Twitter posts and the comments sections of short videos. Using this framework it becomes possible to analyse the differences in digital territorialization projects, different forms of digital mobility and the ways in which socio-spatial positionality matters for the users of digital spaces (Leitner & Sziarto, 2008).
Through this framework, in at least population terms, the United States could be considered to have the largest digital territory in the world, with users from different physical territories around the world congregating in spaces within US digital territory for social, economic, political and cultural reasons. Importantly, the spaces users use are not state-owned but managed by private companies such as Meta, Alphabet and Twitter. 4 The pull of US digital territory is illustrated through the status of UK Facebook users post-Brexit. Upon leaving the European Union (EU), UK Facebook users, no longer covered by EU General Data Protection Regulation (GDPR) regulations that localize their data to the EU, were promptly transferred to a US Terms of Service, with their data moving from the EU to the United States (Menn, 2020). For the United States, while the state is important in the growth of digital territory, including through statesupported digital imaginaries (Graham, 2013), corporate action has been and continues to be essential to the (re)production and expansion of digital territory.
A critical analysis of digital infrastructure can also help to highlight the role of terrainas the political materiality of territory (Elden, 2021, p. 170)in digital territorialization projects. The Pacific Light Cable Network is one particularly clear example of tensions surrounding digital borders, as a cable network originally envisioned to connect California to Hong Kong in the end, due to security concerns from both sides of the digital border, connected the United States to Taiwan and the Philippines (de Seta, 2021). Through this cable network, it becomes evident that the political materiality of digital territory can extend the digital terrain of one nation into the physical territory of other nations (Lim, 2018;Starosielski, 2015), a security and sovereignty quandary that some nations, including the United States, wish to avoid. While sovereignty as a concept operates across multiple scales, digital sovereignty tends to be associated with the capacity of sovereign rulers and authorities (national governments) to exercise supreme power within their territory (Glasze et al., 2022;Kokas, 2022;Pohle & Thiel, 2020). This makes the production of a digital territory and borders one way through which sovereign power can be practiced and extended, as boundaries to sovereign power are a key component of sovereign power. In this manner, territory and sovereignty are intimately connected, with the production of digital territory often one of the first steps to the development of digital sovereignty. With a US state-corporate nexus that has extraterritorial control over digital infrastructure, increased planetary tensions around digital socio-spatial relations are almost inevitable.
Due in part to US dominance over digital relations, the EU has attempted to 'avoid routing data flows within Europe via exchange points and routes outside of Europe' (Pohle & Thiel, 2020, p. 9), with data localization that seeks 'to restrict the storage, movement and/or processing of data to specific areas and jurisdictions' (p. 9) being fundamental to the EU's flagship digital governance policy: GDPR (Chander, 2020;Taylor, 2020). Through GDPR, a legalistic rather than infrastructural intervention, the EU is not just attempting to have control over data abstractly, it is seeking to keep data in a material sphere which it has both physical and digital access to, keeping data on EU terrain. In this manner, data centres that were previously understood as transnational nodes in digital relations become the digital terrain of digital territories, and EU citizens and organizations are asked to carefully consider which terrain they use.
Outside of the United States, China and the EU, it is possible to find evolving digital territorial and sovereignty models, and evolving dynamics within inter-state digital relations (Chen & Yang, 2022). These become particularly visible in Singapore, an important submarine cable and data centre hub. While Singapore has been recently described as one territorial node in evolving geographies of planetary illiberalism (Luger, 2020), the city-state plays a role in ongoing legal disputes elsewhere in the world due to its digital terrain and territory. For instance, a recent legal dispute between the Indian government and Telegram becomes complicated due to Telegram's use of Singaporean data centres, leading one Indian judge to note that 'conventional concepts of territoriality cannot be strictly applied' (The Times of India, 2022). In these examples, whether the language of territory is used (as in India) or avoided (as in the United States), a territorial analysis of digital relations can help pinpoint moments of transnational tension in digital relations (Couture & Toupin, 2019). A wide range of institutions and actors participate in the production of digital territory, terrain and borders in ways that benefit both states and corporations. And while digital territorialization projects are most visible through the actions of large actors, such as the United States, EU, China or India, a number of smaller states and institutions play a vital role in digital territorialization, including through their production of easily accessible digital terrain.
The above examples begin to highlight how a spatial framework for digital relations can shed light on emerging geopolitical tensions. The framework also offers one way through which the geographies of US dominance over digital relations can be mapped and understood relationally (Couldry & Mejias, 2019;Fang, 2018;Fraser, 2019). Importantly for this article, this approach also emphasizes how digital borders take many forms, that they are located somewhere, accept arrivals from multiple places, and have a political materiality and positionality to them. Due to the ways in which digital borders are produced, where the border is and what passes through it often becomes a matter of state and/or corporate policy (Morris, forthcoming, 2023). Through this framework it is possible to highlight the other spatialities involved in digital socio-spatial relations as well as the issues faced by those who have unequal access to digital place, space and mobility.

CHINA'S DIGITAL TERRITORIALIZATION PROJECT
From this conceptual starting point, it is possible to begin an analysis of perhaps the best known digital territorialization project: the Chinese project. The institutions involved in China's digital territorialization project are perhaps the most overt in using the language of space and sovereignty, with the English language name of China's leading body for digital relations being the Cyberspace Administration of China and with the idea of digital sovereignty being used as early as 2010 in the White Paper The Internet in China (State Council Information Office, 2010). But China's digital territorialization project precedes the White Paper by over a decade, with the blocking and filtering of websites and information beginning in the mid-1990s through the creation of a digital border often described as the Great Firewall of China (Griffiths, 2019). This digital border began as a more simplistic filtering project before evolving into a more dynamic system which works within the Golden Shield project, an umbrella project responsible for national digital security that involves the integration and interconnection of previously disparate domestic security systems (Tai, 2010). In recent years this border has been regulated using newer legal technologies, such as the Outbound Data Transfer Security Assessment Measures, which specify what data can and cannot cross digital borders (Cyberspace Administration of China, 2022).
In his book Cyberspace Sovereignty, the lead engineer of this bordering project, Binxing Fang (Fang, 2018, p. 85), suggests that a 'territorial cyberspace' -'cyberspace sovereignty of a state is based on the ICT systems under the state's own jurisdiction'and a 'border' -'the boundaries thereof consist of a collection of the state's own network device ports directly connected to the network devices of other states'are part of a broader Chinese attempt to have sovereignty over the spaces used by Chinese citizens, be they physical or digital spaces, and is in response to US digital power. Fang argues that this sovereignty is based on: the ICT systems under the state's own jurisdiction; the boundaries thereof consist of a collection of the state's own network device ports directly connected to the network devices of other states; cyberspace sovereignty is exercised for protection of various operations of data by cyber roles. The constituting facilities of cyberspace, the carried data and the operation of data are subject to judicial and administrative jurisdiction of the state to which they belong; each state can equally participate in the governance of international network interconnection; operations of the information and communication infrastructure located in the territory of a state shall not be interfered in by other states; a state has the right to protect its own cyberspace from aggression and to maintain corresponding military capabilities. (p. 85) The Chinese digital territorialization project was actualized through the mobilization of engineers (including Fang) and policymakers, with network engineering replacing civil engineering in the production of the border, and new forms of material infrastructure involved in mapping and reinforcing the border (Möllers, 2021;Wilmott, 2020). In this case, the production of a digital terrainincluding servers, cables and other territorialized infrastructurestarted the successful production of a border, and in the following years legal frameworks reinforced this process, aiding in the development of digital sovereignty. In recent years, this has also led to legal frameworks demanding the localization of servers; reproducing digital terrain. In a broader sense, territory here has been a technology used to develop sovereignty over the spaces used by citizens, sovereignty which enables more thorough governance of citizens, including where citizens can congregate, move and what information citizens have access to (Morris, 2022a;Roberts, 2018).
Digital sovereignty is a common reference point in Chinese government documents, either explicitly brought up or alluded to through actions that are only possible due to sovereign control over space. These documents provide a window into a digital spatial imaginary that is extensively used by the Chinese state apparatus. For instance, the 2017 International Strategy of Cooperation on Cyberspace, states: As a basic norm in contemporary international relations, the principle of sovereignty enshrined in the UN Charter covers all aspects of state-to-state relations, which also includes cyberspace. … Upholding sovereignty in cyberspace not only reflects governments' responsibility and right to administer cyberspace in accordance with law, but also enables countries to build platforms for sound interactions among governments, businesses and social groups. … National governments are entitled to administer cyberspace in accordance with law. They exercise jurisdiction over ICT infrastructure, resources and activities within their territories. (Ministry of Foreign Affairs of the People's Republic of China, 2017) Elsewhere, the National Cyberspace Security Strategy notes that 'cyberspace' is as important as 'land, sea, air and space, [that] national sovereignty has extended and stretched into cyberspace, [and that] sovereignty in cyberspace has become an important component part of national sovereignty' (Creemers, 2016). Furthermore, '[n]o infringement of sovereignty … will be tolerated', and as '[c]yberspace is a new territory for national sovereignty' a 'cybersecurity protection [force]' should be built (Creemers, 2016).
Issues of digital sovereignty also appear throughout the Personal Information Protection Law (Creemers & Webster, 2021), with Article 12 noting that China should act as a global norm entrepreneur by: vigorously [participating] in the formulation of international rules [or norms] for personal information protection, [stimulating] international exchange and cooperation in the area of personal information protection, and [promoting] mutual recognition of personal information protection rules [or norms], standards, etc., with other countries, regions, and international organizations.
While in the Data Security Law (Creemers, 2021), the use of Chinese digital terrain becomes a legal requirement for access to China's digital territory, while data mobility to and from Chinese territory also regulated. All these documents are underpinned by the pivotal China's Cybersecurity Law, a law formulated to 'safeguard cyberspace sovereignty and national security' (Creemers et al., 2017).
These documents are part of a legal turn in China's digital territorialization project that began in 2010 and which has accelerated under the Xi Jinping administration. Here, the digital territorialization project evolves from one of sovereign power to one of sovereignty through rule by law (Minzner, 2011). One effect of this legal turn in China's digital territorialization project is that China can now more easily export these ideas, becoming a norm entrepreneur (Creemers, 2020;Erie & Streinz, 2021). But perhaps more importantly for the users of China's physical and digital territory, this legal turn has created a legal process through which companies and citizens can be publicly shamed, rebuked, fined, shut down, imprisoned or intimidated for both their actions in Chinese digital territory and their lack of action in governing the digital spaces that make up China's digital territory, including through official and unofficial measures (Deng & Brien, 2013).
The development of sovereignty through an ever-growing body of legal and government texts means that the Chinese state can govern both the people and the companies that manage the digital spaces that people use, outsourcing governance of the population to private companies that risk being dissolved if they fail to govern correctly. Thus, while in the 2000s China relied on regular infrastructural interventions to produce and govern its digital territory, in the late 2010s and 2020s the Chinese state has been able to rely on legal means to achieve domestically oriented governance goals. This does not mean that the infrastructural interventions have disappeared, but they are no longer the main form of governance, in part because the previous success of infrastructural interventions has led to new governmental challenges emerging.
This legal turn has created an imaginary through which corporations can better understand their responsibilities, users can better understand legal boundaries, and the state can better engage with corporations and users. As I will show below, this understanding of the legal and territorial landscape influences how people use the digital spaces of China's digital territory, including in the practices of contentious politics. In short, the legal turn in China's digital territorialization project has enabled the Chinese state to make 'real the world envisioned by the regulative principles', giving it 'sovereignty over space' (Davis, 2020, p. 51). This has also contributed to planetary geographies of illiberalism, through the exporting of hardware, software and norms (Luger, 2020).

THE SOCIO-SPATIAL EFFECTS OF CHINESE DIGITAL TERRITORY ON #BEIJINGSURGERY#
Scholarship on digitally centred activism and various forms of hybrid activism will be familiar to many readers, and there has been thought-provoking work produced that examines numerous contexts (Castañeda, 2012;Juris, 2012;Recuero et al., 2015;Smit et al., 2017;Su et al., 2022;Yang, 2009). Much of this scholarship has tempered expectations about the power of digital technologies to bring about political change (Chen, 2015;Gerbaudo, 2012). But in these accounts of hybrid and digitally centred activism, the territorial and 'socio-spatial positionality' of those engaged in activism rarely features (Castells, 2015;Leitner & Sziarto, 2008, p. 158).
In much of this scholarship, a relationship between the physical territory of [NATION A] and the digital territory of [NATION B] is present but not interrogated. For instance, during the Arab Spring in Egypt there was an important territorial relationship between Egyptian physical territory and US digital territory, with numerous contentious political projects making use of US digital territory. While this is a somewhat simplistic rendering of the situationother scales do matter and the US digital terrain is selectively spread around the world (Lim, 2018) it gets to a core territorial dynamic that is often alluded to but rarely discussed in ongoing scholarship. Central to this dynamic is the relationship between physical and digital territories that are under different jurisdictions, with the authority governing a physical territory having little control over other digital territories. The result is the use of Twitter, Instagram, Facebook, Telegram, YouTube and Bluetooth mesh networks as sites of activist organization. In many cases this means that those engaged in contentious politics can use spaces that the governments they are protesting against have little control over. As the institution with control over physical territory lacks control over the digital territory it is left with limited options when attempting to govern contentious political action. But with the number of digital territorialization projects increasing alongside attempts to silence activists and state critics through localized and regional internet blackouts (Morris, forthcoming, 2023), it is important to deepen scholarly understanding of the effects that digital territorialization projects have on contentious politics in situations with strong digital borders and advanced digital sovereignty.
To examine this, I return to #BeijingSurgery#, the mapping project described in the introduction. #BeijingSurgery#'s unique trajectory, moving from Weibo hashtag to Weixin instant messaging group project and surviving after initial attempts to end the project through digital governance, makes it a multi-sited activist project. While #BeijingSurgery# did suffer due to deletion and censorship, these forms of governance did not immediately end #BeijingSurgery#, rather, the project evolved. Empirical data from #BeijingSurgery# not only illuminates the domestic spatial effects of a digital territorialization project, but also it shines a light on the role that digital territory and borders play in transnational socio-spatial relations.
#BeijingSurgery# was examined during a multi-year stint of fieldwork in China's hybrid geographies, including the physical spaces of Beijing, the digital spaces of China's digital territories and the spaces outside of China's digital territories that interlocutors visited and congregated in. Under the Xi Jinping administration, the territories in which this study was conducted could be described as spaces of 'authoritarian closure' (Koch, 2013), and this impacted the methods this study could use. To deal with creeping closure, this project made use of a multi-method fieldwork toolkit that included ethnography, semi-structured interviews and playful application walkthrough methods (Morris, 2021b). The ethnographic portions of this fieldwork built on the guidance found in Pink et al. (2015), particularly their principle of decentring the digital in ethnographic work on digital relations. To consider ideas around place, space, positionality and scale the ethnographic practices of Madianou (2016), used to explore co-presence, influenced this study, as did Xiang's (2013) discussion of multi-scalar ethnography.
The Weixin groups discussed in this project were initially accessed through invitation by interlocutors, though these groups were also advertised publicly, through QR codes. These publicly posted QR codes also influenced my understanding of these places, as while I had initially considered these to be private places, I began to see them more as simultaneously semi-public and semi-private; they were accessible to the public, but they were gate kept by administrators; they were visible to censors, but they were private enough to avoid raising an alarm. To me, these places were, in Koch's (2013) words, both 'open' and 'closed', although their level of openness was influenced by the authoritarian nature of the space they were territorialized into.
Once within groups, I made my status as a researcher open knowledge, I included my researcher status in my biography and name, and I negotiated continued access with group administrators. Alongside this I took part in the offline activities of these groups, meeting other group members in person and participating in smaller Weixin groups. During interviews with project leaders and members, I also conducted miniature application walkthroughs (Light et al., 2018), where interlocutors would walk me through their usage of different applications, while also showing me both how they used them and what they could see in them. Through these interactions, it became possible to identify different governance strategies used to control their contentious political action, as well as the tactics different people used to work around forms of governance.

Deletion and displacement
The first obvious effect of China's digital territorialization project on the leaders, members and audience of #BeijingSurgery# was how easy it was to lose access to spaces of protests. This is most obvious through the deletion of the OP and hashtag. After the OP and hashtag were produced, friends of the project leaders and interested others began to repost the OP, the hashtag and the long-image, on Weibo and other digital spaces (Weixin Moments, Douban, etc.). Then, in a moment, and after fewer than 20 hours, #BeijingSurgery#, a complex affective and networked public, was gone (Boyd, 2011;Papacharissi, 2015). The hashtag was deleted, the OP deleted, comments vanished, and many posts sharing the hashtag, resharing the OP or sharing the long-image were shadow banned. This semi-public space, where people came together to discuss the evictions and to share data on them, was gone; the site of resistance had been deleted and the public displaced.
As research on contentious politics and general social media use in China has shown (King et al., 2013(King et al., , 2014, content that is likely to cause protest or that calls for mass gatherings of people is more likely to be deleted than content that simply insults officials. Deletion and censorship of content, information, posts and accounts often occur for political reasons (Chen & Qiu, 2018;Roberts, 2018), and it is more often than not carried out by the operating companies of digital spaces and services in accordance with their legal responsibilities. The Chinese digital territorialization project, the legal turn, and China's digital sovereignty aids the Chinese state in coercing companies into practising deletion; if they do not, they will be exiled, fined or their business dissolved.
But deletion has spatial effects, as it is not an abstract nothingness that is deleted, it is either information (e.g., a news article) or sites where people gather (e.g., hashtags, IM groups, Weibo posts, applications) that are deleted. Focusing on the latter, in the case of #BeijingSurgery#, the deletion of the hashtag and the OP led to the users of these shared digital spaces being displaced, losing access to both their sites of contentious politics and the network of people they met in those and through those places (Morris, 2022a). Space is the central logic of governance here as they lost the site of engagement, not their Weibo accounts. The territorialization project, as a spatial process that enables greater (repressive) governance, was producing spatial effects for those engaged in contentious politics, including forced mobility and displacement.

Semi-private spaces
A secondary effect of #BeijingSurgery#'s displacement was a turn away from semi-public spaces of organization and gathering towards semi-private spaces. My use of semi-public and semi-private here has a double meaning. I am alluding to visibility and ownership, while also considering accessibility. Both Weibo and Weixin are operated by platform companies, companies that market novel affordances alongside the curation of public discourse (Gillespie, 2010). Microblogging spaces such as Weibo are often imagined to offer users the capacity to gather in public and to be visible to other publics, while instant messaging spaces such as Weixin offer the idea of private gathering.
But these spaces are neither fully public nor completely private. First, it is important to remember that no corporate-owned space can truly be public, meaning these digital spaces have much in common with privately owned, semi-public physical places such as London's Canary Wharf and Kings Cross. Therefore, public access to these spaces is always conditional, and one's ability to be visible to others while in them always under negotiation. Furthermore, the 2020 revision of China's Civil Code highlights how digital spaces such as Weixin should be considered semi-private in that they offer privacy from other civil actors but do not offer privacy (through encryption) from the state and its institutional partners, who, under the Personal Information Protection Law, are able to access personal information for reasons of 'security' (Creemers, 2022). These semi-public and private statuses influence what rights users have as well as what personal information is safe from state surveillance: none. Therefore, with the semi-public space of Weibo gone, #BeijingSurgery# became a semi-private, IM group-based project. In using IM groups, #BeijingSurgery# avoided populous semi-public digital spaces where their activities could be easily noticed by a range of people and institutions, instead using IM groups only visible to those within them and the administrators of Weixin.
While the semi-public space was gone, but the project still existed. The #BS# project had included a QR code in the long-image, this QR code was the entry key for the #BS# Weixin IM group. Within hours the IM group had over 99 users, and the QR code stopped functioning; if one scanned it, one would receive an error message informing one that the group one was attempting to join could no longer be joined via the QR code. Despite this form of artificial congestion and gatekeeping the group continued to grow through direct invitation into the group and within 48 hours the group reached close to 500 people, the Tencent mandated size limit of a Weixin IM group. Rather than being affected by a vertical ceiling, the project grew horizontally; soon #BS# IM groups #2, #3 and #4 emerged, alongside administrative groups and data cleaning groups. Each of the numbered #BS# groups had several hundred users, resulting in over one thousand members across several groups.
In this case, China's digital territorialization project resulted in both a change in place of protest, gatekeeping entry into #BeijingSurgery#, a distinct change in the visibility of the project to other, non-member audiences and a change in spatialization, from one large, publicly visible space to numerous smaller spaces. Each of these changes affected #BeijingSurgery# in diverse ways. The change of space meant that the project leaders were now subject to new legislation (China Law Translate, 2017), and the single Group Leader could be prosecuted for content within the group. It also meant that the project went from a leaderless horizontal project to a Group Leader administered IM project.
Gatekeeping through QR code governance meant that entry became invite-only, unlike a hashtag. As one can only invite one's pre-existing Weixin contacts into a Weixin IM group this limited the potential membership of the project, increasing the homogeneity of the project. The QR code gatekeeping and changes in visibility also stopped the project from building on its hashtag momentum, as regardless of how many people saw the hashtag the QR code would not work if more than 99 people were in the group. The move to multiple IM groups worked around the gatekeeping and forced size constraints, but this spatial solution also created problems around information sharing and group consensus. With over 1000 people spread across multiple groups discussions around plans, ideas, analysis and grievances occurred repeatedly, sometimes simultaneously, and it became difficult to share updates across the project's membership. In short, these forms of gatekeeping made it more difficult to extend #BeijingSurgery#'s protest cycle (Tarrow, 1993).

Transnational turn
The third spatial effect of the territorialization project is the use of non-Chinese digital territory for this digitally centred activist project. In #BeijingSurgery#'s case, US digital territory was used through Google Drive, while the more difficult-to-pinpoint abstract non-Chinese digital territories hosting Telegram and Signal were also used. The reasons for this transnational turn were tied to the Chinese territorialization project and the development of digital sovereignty alongside it. As noted above, this gave state and non-state actors a legal requirement to delete and report contentious political activity. The systems of surveillance, censorship and deletion affected what could be said, where people could gather, and what could be stored. Data that were stored in Chinese digital terrainin commercial or personal storagecould be accessed, confiscated and/or deleted by authorities. #BeijingSurgery# lost the data stored in the Weibo hashtag during the initial deletion, but the project also lost data stored in a third-party service, with the service provider informing the project members that their data had to be deleted due to sensitivity. Following these initial deletions, active keyword censorship made it difficult to have sensitive conversations even in semi-private places, and there was always a fear that surveillance technologies would be used to find and prosecute project leaders, members and the Weixin Group Leader. The result was the use of non-Chinese digital territory for project storage and administration, including important administrative conversations, spaces that could be accessed through virtual private networks (VPNs).
Despite the transnational turn in administration and storage, #BeijingSurgery# remained centred in Chinese digital territory. Chinese digital territory was essential for recruiting members, gathering data and accessibility for project members. Not only were many of the non-Chinese digital IM spaces used by project administrators unfamiliar to users of Chinese digital territory, but to access them would require a working VPN, a technology not everybody has access to. For these reasons, while administration and storage were now partially conducted in non-Chinese digital territories, spaces within Chinese digital territory remained vital to the project.
While #BeijingSurgery# did continue to be centred in Chinese digital territory, the project administrators and members made tactical use of non-Chinese digital territory as a spatial imaginary during project moderation. Within #BeijingSurgery# two interlinked spatial imaginaries emerged, the first was that Chinese digital territory was dangerous to people and to the project, an imaginary founded on the deletion of the #BeijingSurgery# hashtag and the recently implemented Internet Group Information Service Management Provisions (China Law Translate, 2017). The second spatial imaginary was that in non-Chinese digital territorythe waiwang 5anything could be discussed; it was a space free from Chinese state censorship and surveillance.
Fearing surveillance and deletion of the #BeijingSurgery# IM groups and an end to the project, project leaders and members used these spatial imaginaries to govern behaviour within #BeijingSurgery#. This was most obvious in how non-Chinese digital territory was described as a space where sensitive conversations that should not happen in Weixin could happen. When language or images were used that were considered dangerous, the group leader, project administrators and project members would suggest these topics be discussed outside China's digital territory, suggesting Telegram, WhatsApp, iMessage, Signal and Wire. Building on a discourse of project safety, group leader safety and project member safety, members may say: Jia: Everybody should really consider using Telegram. Bin: Maybe I am over thinking things though. Jia: Telegram and WhatsApp are both good to use. If you can use them, using them would be smart. 6 These spatial imaginaries had political effects within #BeijingSurgery#, as asking people to move to non-Chinese digital territory for sensitive discussions also silenced those discussions within the group. The silencing strategy of being told to 'discuss this on Telegram' often worked, but there was no infrastructure set up to help these conversations occur in non-Chinese digital territory. While these spatial imaginaries were used successfully, and perhaps productively, what Roberts (2018) describes as 'friction' could stop transnational mobility occurring for the majority of users, meaning that sensitive conversations ended. Overall, the friction to participation created by the digital border meant that #BeijingSurgery# had to remain in Chinese digital territory, and while administrators selectively used (the imaginary built around) other territories, the nature of the project made a full relocation to non-Chinese digital territory difficult and therefore made moderation essential.

Mobility across digital borders
The transnational territorial turn and its associated imaginaries bring me to the final spatiality discussed here, mobility. 7 Digital mobility is associated with all the above examples; forced mobility from hashtags to IM groups; mobility between semi-private IM groups; mobility across transnational digital borders. But another digital border was crossed, taking the digitally centred project into physical spaces and decentring the digital in digital activism (Pink et al., 2015).
China's digital territorialization project, including its associated governance and imaginaries, made moving the project offline at key moments a tactical necessity. #BeijingSurgery# members had key meetings offline, gathered data offline, and engaged in semi-public discussions about practice offline. I caveat this by noting my interlocutors were always simultaneously in a hybrid digital-physical territorythey were always connected and connectingbut, while the plan was for #BeijingSurgery# to be a digitally centred hashtag project, the realities of China's territorialization project meant that interlocutors made an active decision to move across the digital border, to decentre the digital in their digital activism, and to use Beijing's physical territory to avoid digital censorship, deletion and surveillance. While the use of physical spaces in digital activism is not new (Gerbaudo, 2012;Juris, 2012;Lim, 2018), it is often a choice made to extend the scope of the project's activism in an evolving contentious field. For #BeijingSurgery#, mobility across the digital border and decentring the digital was forced on them due to the forms which governance takes in China's digital territory.
Because the #BeijingSurgery# project and its members desired 'stillness' and persistence within Chinese digital territory, hypermobility became a spatial necessity (Bissell & Fuller, 2011;Xiang, 2021). The project had wished to stay in one site, ideally the Weibo hashtag, but what instead emerged was a forced mobility around China's digital territory, Beijing's physical territory and non-Chinese digital territories. The result was a restless, 'turbulent' and 'dynamic stillness' in Weixin IM groups and China's digital territory, a situation where the only way to produce stillness was hypermobility (Martin, 2011;Morris, 2021a;Xiang, 2021).

DISCUSSION: SIMULTANEOUS SITUATEDNESS ACROSS DIGITAL TERRITORIES
The #BeijingSurgery# example shows that digital territorialization projects do not make digital territory static, and digital territorial dynamics and strategies for working around digital sovereignty are important considerations for activists in China. Importantly, in other well-documented examples of digital activism, such as the Arab Spring, the territorial relationship is often bipartite, with a local physical territory and a non-local digital territory. Instead, for #BeijingSur-gery# the core territorial dynamic is tripartite: Chinese digital territory, Chinese physical territory, and non-Chinese digital territory. For #BeijingSurgery# the 'local' digital territory plays a vital role in activist practice. Surveillance and censorship within the bordered digital territory meant that Chinese physical and non-Chinese digital territory became at times safer than Chinese digital territory, even though they were less effective in growing the #BeijingSurgery# project. Thus, #BeijingSurgery# highlights how territorial relations are particularly complex when activists are physically present in territories governed by states engaged in digital territorialization projects.
In contemporary Chinese digitally centred activism, socio-spatial positionality matters in both physical and digital territory (Leitner & Sziarto, 2008). The local physical territory is where the body is present, where the body is in danger, and where the body threatened by eviction and imprisonment exists. Local digital territory and terrain are where the project is spread, where it grows, and, perhaps most importantly, where the everyday lives of project participants take place. The local digital territory is not just a politically contentious field, it is also where family IM groups exist and digital wallets are stored, where sociality occurs, contact are stored, and where likeminded people can be found (Liao, 2019). It may be necessary for professional activities, and it is also a site where 'relational repression' can take place (Deng & Brien, 2013). Chinese digital territory is where activist practice intersects with everyday life, and for those reasons continued use of this territory matters to many activists. To help balance out pressures associated with socio-spatial positionality, #BeijingSurgery#'s leaders and members engaged in tactical simultaneous situatedness across Chinese digital territory, Chinese physical territory, and non-Chinese digital territory.
Being simultaneously situated in multiple territories meant that the project could continue in Chinese digital territory and so could the daily lives of participants. This meant that the most sensitive discussions could take place in either non-Chinese digital territory or Chinese physical territory, away from absolute state surveillance. Being simultaneously situated in three territories at oncea physical territory and two digital territorieswas the case for those in the physical territory of China as well as for the small number of project members based outside of China's physical territory. While those residing abroad did not have the same socio-spatial pressures as some members of #BeijingSurgery#, most of them understood the risks faced by those based in China, particularly the project leaders. Despite the numerous, often repressive spatial effects of China's digital territorialization project, the #BeijingSurgery# project was able to continue by being tactically and simultaneously situated in multiple territories and making use of a range of spatialities.

CONCLUSIONS
In the first sections of this article, I laid out a digital territorial framework based primarily on the work of Elden and Möllers, highlighting how digital territory can be understood as a political technology, how the (re)production of digital territory takes place in part through territorialization projects, the role that digital terrainincluding servers, data centres and cablesplays in this process, and how these ideas help us understand the digital border. In doing this I have shown how digital territorialization projects can take many forms and that they have a multitude of outcomes. I suggested that many people around the world are simultaneously situated in two territories, their local physical territory, and a separate digital territory: often the digital territory of the United States. It is rare, I suggested, for activists to use both local digital and physical territory, a territorial quirk that has important effects on how contentious politics is practiced.
One potential outcome of digital territorialization projects is the use of digital territory as a technology for developing 'sovereignty over space' and making 'real the world envisioned by the regulative principles' (Davis, 2020, p. 51). I examined this outcome through China's digital territorialization project, making use of the writing of Chinese-state affiliated theorist and engineer Binxing Fang (2018), as well as various legal texts, texts that highlight a legal turn in the Chinese digital territorialization project. To understand the form of territory and modes of sovereignty developed during China's digital territorialization project I analysed the digital-activist mapping project, #BeijingSurgery#. During my analysis of #BeijingSurgery#, I highlighted how this project used multiple spatialities in its contentious political practice (Jessop et al., 2008;Leitner & Sziarto, 2008). Notable spatial effects of the Chinese digital territorialization project on #Bei-jingSurgery# were forced mobility through deletion and displacement, the use of semi-private not semi-public spaces for contentious politics, hypermobility between semi-private spaces of contentious politics by project leaders to keep the project on track, the production of multiple small-scale sites of protest, and the use of non-Chinese digital and physical territories to more safely continue the contentious political project. For some leaders and members of #BeijingSur-gery#, this resulted in simultaneous situatedness across multiple digital territories in order to safely continue the project in relation to their socio-spatial positionality.
#BeijingSurgery#'s experience highlights the role of mobility between digital spaces and digital territories within activist practice, with Chinese digital spatial governance requiring regular activist mobility. These forms of cross-border and -territory mobility are discussed in scholarship on activism centred in physical spaces, and the #BeijngSurgery# adds to this literature with a case of digital activist mobility (Castañeda, 2020;Leitner & Sziarto, 2008;Morris, 2021a). Those engaged in #BeijingSurgery# contentious politics were able to find spaces to operate in by exploiting tensions between China's digital territorialization and sovereignty projects and different regimes elsewhere in the world. The Chinese state's focus on macro spatial governance meant that those with alternative spatial imaginaries and realities could effectively practice contentious politics using spatial tactics (Morris, 2022b). This begins to suggest that local digital territories will play a significant role in activist practice despite technologies of digital mobility existing, that both encrypted and public digital spaces are beneficial to activism in China, and that fears of total surveillance may lead to either the digital or the local being decentred in digital activist projects.
#BeijingSurgery#'s experience also forces scholars to question what ideas are normalized in current scholarship on digital activism. For instance, #BeijingSurgery#'s experience suggests that digital space is at least as fragile as physical space, but the fragility of digital space is rarely discussed in current scholarship. In a similar manner, while discussions around digital forgetting seem common, for #BeijingSurgery# being forgotten and erased was all too easy (Ghezzi et al., 2014;Hoskins, 2018). Furthermore, for #BeijingSurgery# the hashtag was a temporary home, not the long-lasting persistent space of activism described in scholarship on anglophone and US digital territory-centred digital activism (Freelon et al., 2018;Juris, 2012). When digital activism is examined from a position centred in US digital territory the question 'What would we do if the projects hashtag was deleted?' is rarely, if ever, asked. Perhaps it should be.
Finally, it is important to remember that digital territorialization projects, as well as the territorial and sovereign models developed in their wake, are bound to neither the US path nor the Chinese path. Just as visa (Wang, 2004), financial (Aalbers, 2018) and political regimes differ across territories, so can digital territorial models. Estonia offers eResidency (Tammpuu & Masso, 2019), Barcelona is imagining new forms of technological sovereignty (Mann et al., 2020), and arguments are being made for digital commons that offer realistic alternatives to some of the digital governance norms being exported by the United States, China and numerous corporations (Arora, 2015;de Rosnay & Stalder, 2020;Fuchs, 2020;Zygmuntowski et al., 2021). In making clearer the potential risks of digital territorial projects due to trends in 'planetary illiberalism' (Luger, 2020), I am not suggesting that states interested in digital territorialization should stop their projects, rather, I hope they consider the potential opportunities for new models of digital territorialization. Neither the United States nor the Chinese model was ideal for the members of #BeijingSurgery#, but that does not mean a territorialization project that could help activist organizations such as #BeijingSurgery# flourish is impossible. I would encourage those engaged in territorialization projects to consider how they can help organizations like #Bei-jingSurgery# to continue their practice safely and effectively while also challenging current regimes of digital mobility and citizenship. around this subject. The author also thanks the anonymous reviewers for their generous comments. The research conducted for this article took place at the London School of Economics; the writing of this article took place at the University of Manchester.

DISCLOSURE STATEMENT
No potential conflict of interest was reported by the author.

ETHICS STATEMENT
All interviewees gave informed consent for interview data to be included in this research project and its subsequent publications. All references to interviewees are anonymised and/or pseudonymised. Data from multiple research participants may be brought together under constructed identities to provide further anonymity to participants in this project. NOTES 1 China's most popular mobile instant messaging service. It is also known as WeChat, which is the international version of Weixin. 2 A long-scrolling image that contains information, shared in image format rather than textual format. It might be created through specific software, or just made through screenshots. 3 A longstanding and extremely popular micro-blogging service. 4 A follow-on effect of this is that academic scholarship also becomes focused on US digital territory. 5 A common phrase in Chinese. Here, wai means external or foreign and wang means the internet. Literally, the internet outside of China's digital border. 6 This is a synthesis of various conversations that took place in #BeijingSurgery# and employs constructed identities. 7 In line with Leitner and Sziarto (2008), I understand 'mobility' and 'socio-spatial positionality' as two of the spatialities composing the multiple spatialities of contentious politics. ORCID Carwyn Morris http://orcid.org/0000-0003-1965-6782